Intelligent Employment Forecasting Model Based On Catboost Algorithm

Currently, there are still many problems in China’s employment situation, among which student employment ability education is the most closely related and direct form of education with economic development. This article intends to use the CatBoost (Categorical Boosting) algorithm to establish a student employment ability prediction model through the processing of category features, automatic processing of missing data, and effective training. Grid search, cross validation and other techniques can be used to select the optimal combination of hyperparameters from the perspectives of learning rate, tree depth, etc. In addition, this article can also address overfitting issues through strategies such as regularization and termination of learning. The data balancing method achieves an accuracy of 0.78, a majority class recall of 0.82, a minority class recall of 0.65, and an F1 score of 0.72 during oversampling. The research results of this article can greatly improve the accuracy of the model in predicting the employment ability of students, and provide scientific data support for the selection of college graduates and employers.